10. Comparison 1 results: subgroup analysis on mortality by wealth quintile and gender.
| Outcome | Subgroup | Trial ID | Study design | Preintervention mortality rate | Postintervention mortality rate | Difference in equity gradient (95% CI) | Analysis summary | ||
| iCCM | Control | iCCM | Control | ||||||
| Change in neonatal mortality rate subgroup (inequity gradient) | Wealth quintile | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | –3.6 (–6.0 to –1.2) | –4.1 (–5.9 to –2.3) |
0.5a (–2.0 to 2.9) P = 0.681 |
Multiple linear regressions adjusted for cluster design and potential confounders |
| Neonatal mortality rate | Wealth quintile (poorest) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 52.1/1000 live births (293/5620) |
54.2/1000 live births (348/6421) | — | |
| Wealth quintile (very poor) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 46.1/1000 live births (248/5380) |
50.2/1000 live births (334/6660) | |||
| Wealth quintile (Poor) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 43.3/1000 live births (252/5818) |
36.0/1000 live births (224/6222) | |||
| Wealth quintile (Less poor) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 39.9/1000 live births (241/6039) |
36.3/1000 live births (218/6001) | |||
| Wealth quintile (Least poor) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 30.9/1000 live births (208/6732) |
33.4/1000 live births (177/5300) | |||
| Change in neonatal mortality rate subgroup (inequity gradient) | Gender | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 1.9 (–4.9 to 8.7) | 2.0 (–3.1 to 7.2) | –0.1a (–8.7 to 8.4) P = 0.974 |
Multiple linear regressions adjusted for cluster design and potential confounders |
| Neonatal mortality rate | Gender (female) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 41.1/1000 live births (557/14,044) |
42.2/1000 live births (614/14,561) | — | |
| Gender (male) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 42.7/1000 live births (667/15,623) |
43.8/1000 live births (712/16,252) | |||
| Change in infant mortality rate subgroup (inequity gradient) | Wealth quintile | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | –2.8 (–4.2 to –1.3) | –4.9 (–7.0 to –2.8) |
2.2a (0 to 4.4) P = 0.053 |
Multiple linear regressions adjusted for cluster design and potential confounders |
| Infant mortality rate | Wealth quintile (poorest) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 38.1/1000 live births (214/5620) |
41.7/1000 live births (268/6421) | — | |
| Wealth quintile (very poor) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 24.9/1000 live births (134/5380) |
32.9/1000 live births (219/6660) | |||
| Wealth quintile (Poor) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 20.5/1000 live births (119/5818) |
24.6/1000 live births (153/6222) | |||
| Wealth quintile (Less poor) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 18.4/1000 live births (111/6039) |
15.2/1000 live births (91/6001) | |||
| Wealth quintile (Least poor) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 14.9/1000 live births (100/6732) |
14.0/1000 live births (74/5300) | |||
| Change in infant mortality rate subgroup (inequity gradient) | Gender | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | –9.1 (–12.2 to –6.0) | –10.8 (–14.7 to –6.9)) |
1.7a (–3.2 to 6.6) P = 0.479 |
Multiple linear regressions adjusted for cluster design and potential confounders |
| Infant mortality rate | Gender (female) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 27.9/1000 live births (392/14,044) |
32.3/1000 live births (471/14,561) | — | |
| Gender (male) | Bhandari 2012a/Taneja 2015 | cRCT | NA | NA | 18.5/1000 live births (289/15,623) |
20.8/1000 live births (338/16,252) | |||
CI: confidence interval; cRCT: cluster‐randomized controlled trial; iCCM: integrated community case management; NA: not applicable. aMultiple linear regressions adjusted for cluster design and potential confounders (distance of nearest point from primary health centre to highway, percent of home births, and years of schooling of mother, gender, religion and caste and wealth quintile).